tion, pooled, and stored at -70'C until further use. BALB/c mice were injected intraperitoneally with 106 specific antibody-producing

Transcription

1 Proc. Natl. Acad. Sci. USA Vol. 75, No. 11, pp , November 1978 Developmental Biology Monoclonal antibody defining a stage-specific mouse embryonic antigen (SSEA-1) (cytotoxicity/immunofluorescence/radioimmunoassay/preimplantation mouse embryo/ stem cells) DAVOR SOLTER AND BARBARA B. KNOWLES The Wistar Institute of Anatomy and Biology, 36th Street at Spruce, Philadelphia, Pennsylvania Communicated by Howard Green, August 17, 1978 ABSTRACT A monoclonal antibody derived by fusion of mouse myeloma cells with spleen cells from a mouse immunized with F9 cells is described. This antibody, which reacts with embryonal carcinoma cells of mouse and human origin and with some preimplantation stage mouse embryos, defines an embryonic stage-specific antigen. This stage-specific antigen (SSEA-1) is first detected on blastomeres of 8-cell stage embryos. Trophectodermal cells are transitorily positive; however, each cell in the inner cell mass eventually expresses this antigen. The importance of embryonic stage-specific molecules in the regulation of cell interactions and cell sorting during development and differentiation has been postulated; however, experimental confirmation of this postulate is scarce. Several candidates for stage-specific molecules have been found by using antisera, raised by both syngeneic (1, 2) and xenogeneic (3) immunization with embryonal carcinoma cells (ECG) or by xenogeneic immunization (4) with mouse embryos. Indeed, inhibition of development of preimplantation mouse embryos by Fab fragments isolated from antisera to the murine cell F9 has recently been reported (5). However, such sera contain antibodies to multiple antigenic determinants, which precludes precise definition of embryo stage-specific antigens. In order to circumvent this difficulty and to study the stage-specific molecules in a methodical fashion, we are producing monoclonal antibodies (6) reactive with cells and embryos. Production and characterization of one such antibody is described here. MATERIALS AND METHODS Preparation of Monoclonal Reagents. BALB/c mice were immunized by weekly intraperitoneal injection of 107 irradiated F9 cells. The mice were tail-bled 7 days after each injection, and sera were tested for reactivity on F9 cells. Three days after the seventh immunization, the spleen was removed from one mouse, which showed a high titer of antibody reactivity. Splenic lymphocytes were isolated and fused with the P3-X63-Ag8 mouse myeloma cell line as described (6-8). Hybrid cell lines were isolated by growth of the fusion mixtures in Dulbecco's modification of Eagle's minimal essential medium containing 10% fetal bovine serum, 2 mm glutamine, and the HAT components (hypoxanthine/aminopterin/thymidine) (9). Supernatants from Linbro wells containing growing colonies were tested for reactivity on F9 cells by an indirect antibody binding radioimmunoassay (RIA). One positive colony was transferred to mass culture and cloned. Supernatants from clones were also tested in RIA and the positive clones were maintained in tissue culture. Supernatants were collected from dense cultures of specific antibody-producing hybrids, clarified by centrifuga- The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U. S. C solely to indicate this fact tion, pooled, and stored at -70'C until further use. BALB/c mice were injected intraperitoneally with 106 specific antibody-producing hybrid cells mixed with Freund's complete adjuvant. Mice with visible ascites were anesthetized and bled by cardiac puncture, and ascites fluid was removed. Sera and ascites fluid were clarified by centrifugation and frozen at -70'C until further use. Cell Lines and Maintenance. For characterization of antibody reactivity, the cell lines listed in Table 1 were used. These cells were routinely grown in Dulbecco's modified Eagle's medium supplemented with 2 mm glutamine and 15% heatinactivated fetal bovine serum without the addition of antibiotics. Embryos. Unfertilized eggs and preimplantation embryos were isolated from random-bred ICR mice as previously described (18). Zonae pellucidae were removed by Pronase digestion and the embryos were left for 4 hr in Whitten's medium before testing. Inner cell masses were isolated from blastocysts by immunosurgery (19). Reagents. Immunoglobulin (Ig) class-specific reagents (goat IgG anti-mouse Au y,782a, y3, K, and X chains) were purchased from Bionetics, Inc. Goat IgG anti-mouse IgM (heavy chainspecific) conjugated with fluorescein isothiocyanate (GAM IgM FITC) and goat IgG anti-mouse IgG (heavy and light chain) (GAMIG) were purchased from Cappell Laboratories, Cochranville, PA. Serum from a rabbit injected with MOPC 104E tumor (Bionetics, Inc.) was utilized as an anti-s chainspecific reagent after passage through a mouse IgG immunoabsorbant column (RAMg). Specificity of all reagents was tested by the Ouchterlony immunodiffusion technique. Reagents were iodinated by the chloramine-t method (20). Rabbit and guinea pig serum were collected, aliquoted, and stored-at -70'C until use as a complement source; these sera were screened for spontaneous cytotoxicity on cells and when necessary absorbed (21) with F9 cells. Techniques. For all serologic tests, cells were harvested by a brief exposure to 0.5% trypsin in 0.1 M EDTA and pipetted to obtain single cell suspensions in 0.01 M Hepes-buffered Eagle's minimal essential medium containing 10% fetal bovine serum. Absorptions were performed either two times with equal volumes of packed cells and diluted antiserum or for quantitative absorption once with known numbers of cells at a constant dilution of antiserum (80% of its maximum binding activity). For all serologic tests, embryos were exposed in 20-1il droplets to increasing dilutions of antiserum in Hepes-buffered Whitten's medium (HWM) (22), washed three times in the same medium, and then transferred to guinea pig complement (1:16 in HWM), GAM IgM FITC (1:10 in HWM), or RAMM labeled with '25I (50,000 input cpm in Dulbecco's modified phosphate-buffered saline with 5% FBS and 0.02% sodium azide) Abbreviations: SSEA-1, stage-specific embryonic antigen 1; ECC, embryonal carcinoma cells; RIA, radioimmunoassay; SV40, simian virus 40; HWM, Hepes-buffered Whitten's medium.

2 5566 Developmental Biology: Solter and Knowles Proc. Natl. Acad. Sci. USA 75 (1978) Table 1. Cell lines and their origins Cell line Species Characteristics Origin SSEA-1* F9 Nullipotent ECC (10) 129/Sv mouse OTT6050 tumor 11,000 PCC4 Pluripotent ECC (11) 129/Sv mouse OTT6050 tumor 10,600 ND-1 Pluripotent ECC (11) 129/Sv mouse OTT6050 tumor 9,400 SCC1 Nullipotent ECC (12) 129/Sv mouse 402-A-II tumor 9,600 NG2 Pluripotent ECC (13) 129/Sv mouse 0TT5584 tumor 10,300 LT/SV Pluripotent ECC LT mouse LT tumor 12,500 MH-15 Nullipotent ECC BALB/c J mouse MH-15 tumor 10,700 FA-25 Nullipotent ECC AKR/J mouse FA-25 tumor 10,400 PYS-2 Parietal yolk sac (14) 129/Sv mouse OTT6050 tumor 0 OTT6050f Fibroblasts 129/Sv mouse 0TT6050 tumor 0 Tera 1 Mixed cultures (15) Human Lung 4,000 containing ECC Tera 2 Mixed cultures (15) Human Lung 3,000 containing ECC Tera 2f Fibroblasts Human Lung 200 B3T3SV SV40-transformed BALB/c mouse Embryo fibroblasts 0 fibroblasts C57SV SV40-transformed C57BL/6J mouse Embryo fibroblasts 0 K129SV SV40-transformed 129/J mouse Adult kidney 0 KGix-SV SV40-transformed 129/J mouse Adult kidney 0 KCA Adenovirus-5-transformed C57BL/6J mouse Adult kidney 0 fibroblasts OAIB Adenovirus-2-transformed C3H/He mouse Adult brain 0 ependymoma BW5147 T-cell lymphoma AKR mouse Thymoma 0 LNSV SV40-transformed Human Skin fibroblasts 700 fibroblasts (17) SV40, simian virus 40. * Stage-specific embryonic antigen 1 measured as cpm bound in indirect 1251 RIA using serum from hybridoma-bearing mouse (dilution 1:32,000). Background cpm is subtracted. For details see legend to Fig. 1. and incubated for 1 hr at 37 C. Cytotoxicity was scored as previously described (18). After six washes in HWM, embryos were transferred to a small drop of the same medium under paraffin oil on a microscope slide and examined with a Leitz microscope equipped with fluorescent epiilumination and X50 water immersion lens. For RIA, embryos were washed six times in phosphate-buffered saline/fetal bovine serum/azide, transferred to tubes, and assayed for radioactivity in a Packard gamma counter. RIA and antibody-mediated complementdependent lysis on tissue-cultured cells were performed as previously described (23). RESULTS Detection and Isolation of Antibody-Producing Hybrid Cells. Supernatant fluids from 14 separate colonies of cells proliferating in hypoxanthine/aminopterin/thymidine medium were tested in RIA on target F9 cells with 125I-labeled GAMIG as the second reagent. Supernatant from one of these colonies was found positive. This colony was propagated in vitro and cloned by limiting dilution. Characterization of Antibody. Supernatant fluids from the parental colony and a clone were tested with mouse immunoglobulin class-specific reagents in Ouchterlony double immunodiffusion. Precipitation lines of identity were found with goat IgG anti-mouse 'y (secreted by the parental myeloma), with goat IgG anti-mouse g chain sera, and with goat IgG anti-mouse K; no activity with the anti-mouse X chain sera was found. Sodium dodecyl sulfate gel electrophoresis of reduced immunoprecipitates (using RAMu) of radiolabeled supernatants from the hybrid colony showed a Coomassie blue-stained band of 68,000 Mr, compatible with the molecular weight of mouse,u chain. After these determinations were made, anti-mouse,u chain-specific reagents were used in all indirect immunologic testing. Sera and ascites fluid from hybrid tumor-bearing mice were titered on F9 target cells. In the interests of uniformity the serum from one of these mice has been used for the remainder of the work reported herein. Reactivity of the Monoclonal Reagent on Cell Lines. In order to determine specificity of the monoclonal reagent, it was reacted on a series of mouse and human cell lines in RIA (Table 1 and ref. 24). The reagent was titered on F9 target cells and the remaining cells were tested on the plateau of the titration curve (1:8000) and with two dilutions (1:32,000 and 1:128,000) in the descending portion of curve. The reactive cells include each mouse stem cell line tested, and two human -derived cell lines. Nonreactive cells include several nonstem cell lines derived from mouse and human and several mouse transformed cell lines derived from different inbred mouse strains. These results were corroborated by absorption of the monoclonal reagent with the same cell lines. When quantitative absorption was performed with increasing numbers of cells at 1:32,000 dilution (approximately 75% of maximum binding on F9 cells) and the absorbed reagent was tested on F9 cells, the cell lines again fell into the same distinct categories (Fig. 1). Each of the mouse stem cell lines tested absorbed the antiserum reactivity; MH was most effective (5.5 X 104 cells removed 50% of the binding activity on F9 cells) and

3 6- ~~~~~~~~~~~~~Te ra 1 x 5 - E Developmental Biology: Solter and Knowles ~~~~~~~~~NG2 4- ~~~~~~~~PCC4 2-FA 1 MH Cells FIG. 1. Absorption analysis of antibody reactivity by RIA on F9 target cells. Aliquots (200,gl) of serum diluted 1:32,000 were absorbed with 106, 105, and 104 of each of the cell lines (40C, 1 hr). The absorbed sera were clarified by centrifugation and tested in indirect RIA in triplicate on F9 target cells. Triplicate replicas of 5 x 105 cells were incubated with 100,l of diluted supernatant (4 C, 1 hr); the final serum dilution was 1:64,000 (50,ul absorbed serum added to 50 Jl containing 5 X 105 cells). The cells were washed three times in phosphate-buffered saline/5% fetal bovine serum/0.02% azide and then incubated with 100 Ml of 125I-labeled RAMg diluted to give 50,000 input cpm. After an additional three washes, the cells were transferred to tubes for gamma counting. As a control, F9 was incubated with the same dilutions of normal BALB/c serum and the average cpm ( ) was subtracted from the total cpm bound to give specific cpm. Cells that did not absorb any activity (shaded area) include K129SV, KGix-SV, B3T3SV, C57SV, KCA, OAIB, OTT6050f, PYS, LNSV, BW5147, and Tera 2f. Proc. Natl. Acad. Sci. USA 75 (1978) 5567 Table 2. Complement-mediated antibody-dependent cytolysis Dilution of antiserum togive 50% lysis Cell lines >1:800,000 F9, MH, SCCI 1:140,000 ND-1 1:100,000 PCC3, PCC4 <1:100 PYS, B3T3SV, OAIB, Tera 2f, Tera 1* * At dilutions up to 1:10,000, approximately 20% specific lysis was observed NG-2 was the least active (8 X 105 cells were required to remove the same amount of antiserum reactivity). With the human cell lines, 40% of the reactivity was removed by 106 Tera 1 cells, and 106 Tera 2 cells removed 20% of the binding activity on F9 cells. The binding activity of supernatant from the hybridoma was also tested after a single absorption at 1:10 dilution with 1.2 X 108 epididymal mouse sperm (data not shown), and it no longer reacted with F9 target cells. When volumes of packed cells equivalent to the packed cell volume of 106 F9 cells were used to absorb 200,ul of 1:32,000 diluted antisera human sperm, 129 strain mouse kidney and brain removed 80% of the binding ability of the monoclonal reagent on F9 cells while an equivalent volume of mouse spleen cells did not affect the binding. Culture fluid from growing hybridomas and serum from hybridoma-bearing mice were titered in complement dependent lysis. The dilution of supernatant from the antibodyproducing hybrid cells necessary to lyse 50% of F9 cells was 1:50, while serum from hybrid tumor-bearing mice lysed 75% of F9 cells at 1:800,000 dilution (Table 2). Other embryo-derived murine cells were also lysed by this serum and complement with high efficiency. A small percentage of the Tera 1 human -derived cell lines were lysed at low dilutions of serum. The human cell line Tera 2 and control mouse and human cell lines were not lysed (Table 2). Reactivity of the Monoclonal Reagent with Preimplantation Mouse Embryos. Embryos were tested with supernatant fluids from the hybridoma culture and with serum from hybrid tumor-bearing mice in RIA. No antibody was bound to unfertilized eggs, zygotes, and 2- and 4-cell stage embryos. Late 8-cell stage embryos and morulae bound antibody with increasing efficiency, and the amount of binding was decreased on blastocysts. An increase in the amount of antibody bound was seen on inner cell masses, both freshly isolated and grown in culture, compared with blastocysts (Fig. 2). These same embryonic stages were tested in complementdependent lysis. No lysis of embryos prior to the 8-cell stage was observed. Ten to 20% of 8-cell stage embryos were lysed with antiserum dilutions up to 128,000. Morulae, blastocysts, and inner cell masses were lysed with high efficiency (Table 3). Similar results were obtained with embryos in indirect immunofluorescence assays; unfertilized eggs, zygotes, and 2- and 4-cell stage embryos were negative. Of the 8-cell stage embryos investigated, approximately 10% consisted of all positive blastomeres, 10-20% had some positive and some negative blas E.. a L - 3 e t s t C >N N s X O C) N N 000o Go ~0 o ~~~~~~o0 FIG. 2. Indirect 1251-binding assay of serum from hybridomabearing mouse to preimplantation stage mouse embryos. Ten embryos of each stage were incubated with antiserum diluted 1:100 or normal BALB/c serum (1:10) as a control. The graph represents the average value from five separate experiments. ICM, inner cell masses. Numbers after development stages denote hours of in vitro cultivation prior to analysis.

4 5568 Developmental Biology: Solter and Knowles Table 3. Complement-mediated lysis of preimplantation mouse embryos by monoclonal antiserum* Developmental Dilution of antiserum stage to give 50% lysis Unfertilized egg Zygote <10 2-cell stage j <15 4-cell stage J 8-cell stage Morula Blastocyst Inner cell masses nt experiments; 10 embryos used for each dilution of antiserum in e-ach experiment. embryos. Inner cell masses grown in vitro for up to 3 days were positive, although some negative cells were also present. Ectoderm exposed by removing the outer layer of entoderm from inner cell masses grown in vitro (25) was always completely positive (Fig. 3g). DISCUSSION The monoclonal antibody described here is specific for an an- <1: :256,000 tigenic determinant present on stem cells, be they of human or mouse origin. Allother differentiated cell lines 1:64,000 tested were negative, including differentiated cell lines derived 1:64,000 from the same tumors as the stem cell lines. * Cumulative data of six independei Because some of the mouse and human cell lines are mixed cultures containing both differentiated and t Lysis was not observed at any diluition. undifferentiated cells, it is possible by varying the culture Lysis of 10-20% of embryos was observed at end dilution up to conditions to greatly decrease or completely eliminate stem cells 1:128,000. f Even at lowest dilutions, 10-30% c embryos were not lysed. from some cultures as judged by morphological criteria. Such depleted cultures reacted very poorly or not at all with mono- were totally negative. The clonal antibody (data not shown) in RIA and in immunofluo- tomeres, and the remaining embrryos rescence. difference between positive and r iegative.blastorneres was easily discernible (Fig. 3) with antiseruur dilutions b :to of 0. At later Appearance of antigen at late 8-cell stage and its disap- inner cell masses) pearance from trophectoderm might indicate the beginning developmental stages (morulae, Iblastocysts, the number of completely and Ipartially positive embryos in- and the cessation of synthesis or the uncovering and covering creased significantly (70-90%); hkowever, throughout the range of the existing antigen. However, the reason for the unequal of antiserum dilutions there weree always some totally negative _ W - al distribution of stage-specific embryonic antigen 1 (SSEA-1) seen on preimplantation embryos is not clear. Because no consistent pattern of positive cells could be discerned, it is unlikely that the antigen represents a marker of any cell lineage unless we assume that cells with different developmental fate are randomly distributed in mouse morulae. The other possibility is that antigen is expressed differently during the cell cycle, as has been observed with other antigens. SSEA-1 is present in most but not all cells of freshly isolated inner cell masses. When inner cell masses are grown in vitro, they develop a layer of entodermal cells on the outside (19), and this layer is mostly positive after 48 hr of culture and remains so for all the time we were able to examine it (120 hr). The entodermal layer can be selectively removed by repeated immunosurgery (25), and thus the exposed ectodermal layer uniformly positive throughout the entire culture period ( was hr) Ṡeveral conventional antisera raised by syngeneic and xenogeneic immunization against stem cells and preimplantation mouse embryos have been described (for review see refs. 1, 3, 26, 27). In an attempt to compare our monoclonal reagents with these previously described antisera, we have to bear in mind that the monoclonal antibody might represent a part of the activity of any of the conventional antiserum. This may be the case when a conventional antiserum reacts with the same cells and preimplantation stage embryos as the monoclonal reagent, even though the conventional antiserum may react with additional cells. However, if the monoclonal reagent reacts with cells or embryonic stages with which a conventional antiserum does not react, then we can conclude that monoclonal reagent is not part of activity in the conventional antiserum. On the basis of these premises, antisera to the F9 antigen (28) do not contain reactivity to the same antigenic specificity as the monoclonal reagent because the activity of F9 antiserum is not removed by absorption with brain and kidney. On the other hand, absorption with tissue homogenates is difficult to control and we cannot exclude the possibility of nonspecific absorption with brain and kidney. The ); (c) 8-cell stage, four positive anti serum described by Stern et al. (2) has a FIG. 3. Preimplantion mouse embryos were exposed to serum from a hybridoma-bearing mouse for 1 hr, washed, and exposed to GAM IgM-FITC for 1 hr. (X300.) ( Ia) Four-cell stage; (b) 8-cell stage, twq. positive blastomeres (arrows blastomeres; (d) morula, two positlive blastomeres (arrows); (e) mor rula, completely positive; (f) blastbocyst, weakly positive mural tro- range of activity very similar to that of our monoclonal antibody phectoderm; (g) ectoderm from inner cell mass grown in vitro for 3 and its activity is also absorbed with mouse brain and days-all cells positive. kidney. Proc. Nati. Acad. Sci. USA 75 (1,978)

5 Developmental Biology: Solter and Knowles Similarly, antiserum against antigen I (3) defined by rabbit antiserum to mouse cells possessed, in addition to other activities, all the activities displayed by this monoclonal antibody. Additionally, in parallel with our results, antigen I antiserum does not react with trophectodermal cells of the late, hatched blastocyst. The monoclonal antibody described herein can thus be a part of several described antisera, and confirmation of this must await the molecular identification of the antigenic specificity. Our preliminary data from immunoprecipitation of extracts of 125I-labeled F9 cells and sodium dodecyl sulfate gel electrophoresis indicate that SSEA-1 might be a glycolipid because specifically precipitated material runs in front of bromophenol blue marker and is extractable with chloroform/methanol. During preparation of this manuscript we became aware of recent results obtained by Stern et al. (29) and Willison and Stern (30). They hybridized mouse myeloma cells with spleen cells from rats immunized with mouse spleens and isolated antibody-producing hybrids. One monoclonal antibody reacted with mouse cells and sheep erythrocytes. Detailed analysis of this antibody indicated that its reactivity pattern coincides with the distribution of Forssman antigen. The reactivity of this antiserum on cells and preimplantation mouse embryos closely resembles the activity of our monoclonal antibody. We, therefore, tested our monoclonal antibody against sheep and ox erythrocytes, using hemagglutination, hemolysis, and RIA as well as absorption and testing of absorbed serum on F9 cells (data not shown). No reactivity of the monoclonal antibody to SSEA-1 was found with sheep erythrocytes in any direct tests and absorption with both sheep and ox erythrocytes removed the same minimal amount of activity. We can conclude that SSEA-1 is different from Forssman antigen, though it appears that both might be present on the same cells and embryonic stages. Because the Forssman antigen is also a glycolipid, it is possible that several glycosyltransferases are synthesized or activated during early preimplantation development and their products are then presented on the embryonic cell surface. SSEA-I begins to be expressed on 8-cell stage embryos and the Forssman antigen on late morulae. Concordant biochemical analysis of the antigenic specificities and reactivity patterns on early mouse embryos should elucidate some of the cell surface changes during early development. This work was supported in part by the National Cancer Institute (CA-18470, CA-17546, CA-10815, CA-21069), the American Cancer Society (PDT-26, IM-88), The National Foundation-March of Dimes (1-301), and National Institute of Allergy and Infectious Diseases Research Career Award to B.B.K. (AI-00053). We thank Peter Stern for preprints of his work and Karen Artzt, Carlo Croce, Hedwig Jakob, Gail Martin, Cesare Milstein, and Steven Pfeiffer for cell lines. The technical assistance of K. Maloney, L. Shevinsky, H. Wang, R. Lin, and J. Barnes is gratefully acknowledged. We wish to thank David Aden for his discussions and support throughout this work. Proc. Natl. Acad. Sci. USA 75 (1978) Jacob, F. (1977) Immunol. Rev. 33, Stern, P. L., Martin, G. R. & Evans, M. J. (1975) Cell 6, Edidin, M. & Gooding, L. R. (1975) in Teratomas and Differentiation, eds. Sherman, M. I. & Solter, D. (Academic, New York), pp Willey, L. M. & Calarco, P. G. (1975) Dev. Biol. 47, Kemler, R., Babinet, C., Eisen, H. & Jacob, F. (1977) Proc. Natl. Acad. Sci. USA 74, Galfre, G., Howe, S. C., Milstein, L., Butcher, G. W. & Howard, J. C. (1977) Nature (London) 266, Kohler, G. & Milstein, L. (1975) Nature (London) 256, Kohler, G. & Milstein, L. (1976) Eur. J. Immunol. 6, Littlefield, J. W. (1964) Science 145, Bernstine, E. G., Hooper, M. L., Grandchamp, S. & Ephrussi, B. (1973) Proc. Natl. Acad. Sci. USA 70, Jakob, H., Boon, T., Gaillard, J., Nicolas, J. F. & Jacobs, F. (1973) Ann. Microbiol. (Paris) 124B, Martin, G. R. & Evans, M. J. (1975) Cell 6, Dewey, M. J., Martin, D. W., Martin, G. R. & Mintz, B. (1977) Proc. Natl. Acad. Sci. USA 74, Lehman, J. M., Speers, W. C., Swartzendruber, D. E. & Pierce, G. B. (1974) J. Cell. Physiol. 84, Fogh, J. & Trempe, G. (1975) in Human Tumor Cells in Vitro, ed. Fogh, J. (Plenum, New York), pp Knowles, B. B., Ford, S. R. & Aden, D. P. (1977) J. Immunol. 119, Croce, C. M., Girardi, A. J. & Koprowski, H. (1973) Proc. Natl. Acad. Sci. USA 70, Solter, D. & Schachner, M. (1976) Dev. Biol. 52, Solter, D. & Knowles, B. B. (1975) Proc. Natl. Acad. Sci. USA 72, Hunter, W. M. (1973) in Immunochemistry, ed. Weir, D. M. (Blackwell, Oxford), pp Goldberg, E. H., Boyse, E. A., Bennett, D., Scheid, M. & Carswell, E. A. (1971) Nature (London) 232, Whitten, W. K. (1970) Adv. Biosci. 6, Knowles, B. B., Solter, D., Trinchieri, G., Maloney, K. M., Ford, S. R. & Aden, D. P. (1977) J. Exp. Med. 145, Knowles, B. B., Aden, D. P. & Solter, D. (1978) in Current Topics in Microbiology and Immunology, eds. Melchers, F., Potter, M. & Warner, N. (Springer, Heidelberg, W. Germany), Vol. 81, pp Strickland, S., Reich, E. & Sherman, M. I. (1976) Cell, 9, Graham, C. F. (1977) in Concepts in Mammalian Embryogenesis, ed. Sherman, M. I. (MIT Press, Cambridge, MA), pp Solter, D. (1977) in Immunobiology of Gametes, eds. Edidin, M. & Johnson, M. H. (Cambridge University Press, London), pp Artzt, K., Bennett, D. & Jacob, F. (1974) Proc. Natl. Acad. Sci. USA 71, Stern, P., Willison, K., Lennox, E., Galfre, G., Milstein, C., Secher, D., Ziegler, A. & Springer, T. (1978) Cell 14, Willison, K. R. & Stem, P. L. (1978) Cell 14,